© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Feature Splitting for GPT2-Small
    5. GPT2-Small
    7. Residual Feature Splitting 24576
    9. 8-RES_FS24576-JB
    11. 4472
    Prev
    Next
    INDEX
    Explanations

    repetitive phrases or actions, particularly related to beliefs and actions that lead to negative cycles

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs24576-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    Data Type
    torch.float32
    Hook Point
    blocks.8.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    8
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    373
    +0.10
    0.5%
    340
    +0.08
    0.4%
    512
    +0.08
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    290
    +0.10
    0.02
    340
    +0.08
    0.02
    716
    +0.08
    0.02
    Negative Logits
     warr
    -0.59
    {"
    -0.55
    nostic
    -0.55
     Dynam
    -0.54
    Grey
    -0.54
    Dou
    -0.54
    iciary
    -0.53
     Guid
    -0.52
     Fighter
    -0.52
    leigh
    -0.51
    POSITIVE LOGITS
    âķIJ
    0.82
     parcel
    0.76
    rogen
    0.72
    rogens
    0.72
    urger
    0.70
     beyond
    0.69
    hra
    0.64
     forget
    0.62
    =-=-
    0.62
    200000
    0.61
    Activations Density 0.044%

    No Known Activations